In the design and construction of systems for high power radio frequency (RF) generation, it is desirable to develop systems with improved conversion efficiencies, as the operating temperatures of the component power amplifier (PA) transistors are often a limiting factor. A variety of circuits have been developed in order to increase the efficiency of an RF power amplifier by exploiting the switching characteristics of a power amplifier device. Such systems are known in the art as switching power amplifiers. The most common switching power amplifiers include the class-D, class-E, and class-F switching circuits.
As shown in FIG. 1, a class-D power amplifier includes a first switching device Q and a second switching device /Q arranged in a “totem pole” configuration. Each device is driven to conduct complementarily through 180 degrees out of each RF cycle. The output of a class-D power amplifier is a square-wave voltage waveform, with the amplifier being fed from a low-impedance voltage bus. The load network of the class-D power amplifier includes a series-tuned band pass filter, permitting the conduction of a current waveform consisting primarily of a carrier frequency having highly attenuated harmonics. As a result, the current conducted by each PA switch is a half-sinusoid waveform.
Those skilled in the art will recognize that traditional power switching devices suffer from parasitic reactances. The parasitic reactances exist due to capacitance associated with the drain-source of a component PA transistor and the inductance associated with the transistor interconnections. Parasitic reactances may cause undesired resonances within the PA transistor waveforms, and may degrade efficiency at higher frequencies.
FIG. 2 illustrates a schematic view for a class-F power amplifier utilized to ameliorate the effects of the parasitic reactances of the transistors. The class-F amplifier consists of a choke-fed, grounded-source switching device, which is connected via a quarter-wave transmission line (labeled ¼ wave in FIG. 2) to a parallel LRC band-pass filter. The band-pass filter of the class-F PA presents a low-magnitude impedance to all harmonics except the fundamental of the frequency of operation. The ¼ wave transmission transforms the impedance in a manner which presents the drain of the transistor with a high impedance to all odd-order harmonics. Consequently, the drain voltage waveform is a square wave, while the drain current waveform is a half-sinusoid, conducted during the time frame in which the drain-source voltage is at its minimum. Moreover, the transistor interconnection inductance may be considered absorbed into the transmission line section.
FIG. 3 illustrates a schematic view of a push-pull class-F amplifier, which is a variation of the class-F power amplifier depicted in FIG. 2. The push-pull class-F power amplifier may be constructed by replacing the RF choke of the tradition class-F power amplifier with an additional switch and transmission line. The second transistor of the push-pull class-F power amplifier is driven in opposition to the lower device, creating a push-pull circuit.
FIG. 4 illustrates a class-E power amplifier. The class-E power amplifier topology is capable of absorbing both the transistor parasitic shunt capacitance and the interconnect inductance into its resonant circuit. The output network of the class-E power amplifier has a net impedance at the frequency of operation with a nominally 52 degree inductive component, while presenting a much larger impedance at all higher harmonics. The switch of the class-E power amplifier is driven at a nominal 50% duty, with the drain-source voltage waveform being a damped sinusoid, which returns to zero volts prior to the commencement of the conductive half-cycle of the switch. Moreover, the drain current of the class-E power amplifier is a sinusoidal segment, beginning at zero amps at the commencement of the conductive half-cycle.
FIG. 5 illustrates a simplified version of the class-E power amplifier circuit. In the simplified class-E power amplifier circuit, the output loading circuit is placed between the supply voltage and the resonant switch circuit.
The combination of power amplifiers, such as class-D, class-E, and class-F power amplifiers, poses several difficulties. Power amplifiers, unlike passive components, cannot simply be connected in parallel or series to sum their outputs. It is therefore desirable to create a circuit or system of circuits which effectively combine various power amplifiers. More specifically, it is desirable to create a distributed amplifier utilizing a number of switching power amplifier circuits.